MBB Nerd Polls 2024-25

I had some time and inclination to look up the pre-tourney stats from KenPom. As Kedsey noted, they only go back to 2001. Here are all the teams that were in the top 5 of both AdjOE and AdjDE, 2001-2024:

YearTeamAdjOEAdjDENCAAT
2001Duke22National Champions
2002Duke12Sweet 16
2004Duke33Final Four
2005Illinois25National Runner-Up
2007North Carolina32Elite 8
2008Kansas13National Champions
2010Kansas24Round of 32
2010Duke45National Champions
2019Virginia25National Champions

From this group...
100% Reached the Round of 64
100% Reached the Round of 32
88.9% Reached the Sweet 16
77.8% Reached the Elite 8
66.7% Reached the Final 4
55.6% Reached the National Title Game
44.4% Won the National Championship

Those are pretty good outcomes if you ask me.

There is a larger group of teams that were Top 10 in AdjOE and AdjDE. Here's that group, 2001-2024 (excluding Kansas in 2020 due to the lack of an NCAA Tournament that year):

YearTeamAdjOEAdjDENCAAT
2001Duke22National Champions
2001MSU38Final Four
2001Arizona103National Runner-Up
2002Duke12Sweet 16
2002Cincinnati71Round of 32
2002Kansas57Final Four
2003Kentucky84Elite 8
2004Duke33Final Four
2005Illinois25National Runner-Up
2005North Carolina36National Champions
2005Duke103Sweet 16
2006UConn89Elite 8
2007North Carolina32Elite 8
2007Ohio State610National Runner-Up
2008Kansas13National Champions
2008UCLA64Final Four
2008Duke87Round of 32
2010Kansas24Round of 32
2010Duke45National Champions
2011Duke65Sweet 16
2012Kentucky26National Champions
2012Ohio State73Final Four
2013Florida64Elite 8
2014Louisville76Sweet 16
2015Kentucky61Final Four
2016Kansas74Elite 8
2016Virginia86Elite 8
2017Gonzaga102National Runner-Up
2018Duke37Elite 8
2018Michigan99National Runner-Up
2019Virginia25National Champions
2019Duke66Elite 8
2019Michigan State48Final Four
2019North Carolina710Sweet 16
2021Gonzaga110National Runner-Up
2021Illinois75Round of 32
2021Michigan67Elite 8
2022Gonzaga17Sweet 16
2024Auburn104Round of 64

From this group...
100% Reached the Round of 64
97.4% Reached the Round of 32
87.2% Reached the Sweet 16
71.8% Reached the Elite 8
48.7% Reached the Final 4
30.8% Reached the National Title Game
15.4% Won the National Championship

At the moment, Duke and Iowa State are the only two teams in top 10 of both AdjOE and AdjDE.
While it's difficult to predict the future I'd wager a pie bet against both Duke and Iowa State winning the national championship.
 
After Duke's low scoring win over Wake, our KenPom offensive efficiency dropped to #6, but with Houston's high scoring game with Kansas, we will probably take over #1 on defensive efficiency.
 
It would be fun to pop ahead of Auburn in the NET tomorrow. No way to know if we are close.
I guess it's "fun" to be #1 in NET, but meaningless even for NCAA tournament seeding purposes. Just helps Kentucky and Kansas who then have a win over the #1 team, but ironically, doesn't help us being #1 vs. #5 or whatever because we're just judged on the quality/quantity of our wins/losses. It's used as the blunt measurement to assess the quality of your opponents and then you're judged based on how you do against that. Of course, there's high correlation between raw NET rank and results...
 
I guess it's "fun" to be #1 in NET, but meaningless even for NCAA tournament seeding purposes.
One consequence of getting the overall #1 seed would be getting to choose our region (I assume that's still a thing?) Of course, I don't even know if we would choose the East (Prudential Center in NJ) or the South (State Farm Arena in Atlanta) but it would be nice to have the choice.
 
I guess it's "fun" to be #1 in NET, but meaningless even for NCAA tournament seeding purposes. Just helps Kentucky and Kansas who then have a win over the #1 team, but ironically, doesn't help us being #1 vs. #5 or whatever because we're just judged on the quality/quantity of our wins/losses. It's used as the blunt measurement to assess the quality of your opponents and then you're judged based on how you do against that. Of course, there's high correlation between raw NET rank and results...
The official stance of the NCAA is that '"the NET is one of many tools used by the committee in seeding the tournament". So it's not the final answer - quadrant wins and all matter - but it's a factor. Would be interesting to see how many of the NET #1s were the overall tournament #1 seed the past few years. I would guess there's a strong correlation.

Does it matter this year so much for Duke? Probably not. As long as we are a top 3 seed, we very likely get #1 in the East. It might impact, for example, whether we get the #4 or #5 SEC team as our #2 seed. But that's overthinking it. We just need to keep winning and get that East #1 seed.
 
The official stance of the NCAA is that '"the NET is one of many tools used by the committee in seeding the tournament". So it's not the final answer - quadrant wins and all matter - but it's a factor. Would be interesting to see how many of the NET #1s were the overall tournament #1 seed the past few years. I would guess there's a strong correlation.

Does it matter this year so much for Duke? Probably not. As long as we are a top 3 seed, we very likely get #1 in the East. It might impact, for example, whether we get the #4 or #5 SEC team as our #2 seed. But that's overthinking it. We just need to keep winning and get that East #1 seed.
"Who did you beat and where did you beat them?" seems to be the committee's seeding mantra, though NET rankings near the top are going to mostly capture that.

Looking at last year's Selection Sunday NET, the two teams from the top two seed lines whose NET-to-seed disparity stand out are North Carolina (shock of shocks) and Marquette. UNC was only No. 8 in the final NET ranking, but was seeded No. 4 overall; Marquette was No. 14 in NET and seeded No. 7 overall.

It gets messier from there though. Looking at 3 seeds, Baylor was No. 15 in NET (incl. a neutral loss to No. 10 Duke), yet seeded No. 9 overall while Kentucky was No. 18 NET and No. 11 tourney.

Meanwhile, Alabama, Duke and BYU were 9, 10 & 12 in NET yet seeded overall Nos. 16, 13 & 17. All potential/likely 3 seeds who were seeded 4/5 instead.

Each of the past two years, Houston was No. 1 in NET and the team just behind them got the overall No. 1 seed. In 2022, Gonzaga and Arizona were 1-2 in both (while poor Houston was No. 3 in NET and given the No. 18 overall seed).
 
Over the first 12 possessions of the second half against Wake (about 8 minutes, or 20% of the game), Duke was outscored 17-1. And yet, the team played well enough during the rest of the game to manage a 95 Torvik game score. Which got me thinking.

Most of you know that Torvik calculates a "game score" for each game a team plays, with a maximum of 100 being the best a team can play. Every year, no matter how good the Duke team is, the team inevitably has a couple of "clunkers" (which for these purposes I've defined as a game score of 75 or worse). This eight minute segment against Wake represents as close to a clunker as the 2025 Blue Devils have come, but we've only played 19 games. So I compared our game score performance against the first 19 games of past Duke teams (back to 2008, which is the earliest Torvik has on his website):

YearAvgMedian<90<=75Low<=75 in games 20+NCAAT
2025​
95.89​
97​
1​
0​
87​
????
2008​
95.84​
98​
3​
0​
81​
2
16​
2009​
95.32​
98​
2​
1​
73​
3
16​
2019​
95.21​
98​
3​
0​
79​
2*
8​
2011​
94.84​
96​
2​
0​
80​
2
16​
2010​
93.58​
97​
3​
1​
48​
1
1​
2015​
91.58​
97​
5​
2​
48​
2
1​
2014​
91.42​
96​
5​
2​
70​
2
64​
2013​
91.26​
96​
3​
1​
29​
3
8​
2012​
90.47​
92​
7​
1​
59​
4
64​
2016​
90.00​
93​
7​
1​
75​
2
16​
2018​
89.95​
95​
7​
4​
63​
1
8​
2020​
89.89​
97​
7​
3​
64​
3n/a
2017​
89.63​
93​
5​
2​
40​
1
32​
2022​
89.58​
90​
7​
2​
68​
2
4​
2024​
83.11​
92​
8​
5​
40​
3
8​
2023​
78.47​
88​
11​
4​
9​
5
32​
2021​
75.79​
76​
11​
9​
50​
2n/a

* (both of the 2019 team's clunkers came while Zion was injured)

As you can see, this year's model has the best average game score, the fewest games under 90, the highest low score, and is one of only four Duke teams in the period to have no clunkers. For the first 19 games anyway. That shows not only how high a level we've played so far, but how consistent we've been (notwithstanding the first 8 minutes of the second half against Wake).

That said, Duke has had at least two clunkers from the 20th game and beyond in 14 of the previous 17 seasons (all but 2010, 2017, and 2018). Can we avoid that this year? Only time can tell.
 
Over the first 12 possessions of the second half against Wake (about 8 minutes, or 20% of the game), Duke was outscored 17-1. And yet, the team played well enough during the rest of the game to manage a 95 Torvik game score. Which got me thinking.

Most of you know that Torvik calculates a "game score" for each game a team plays, with a maximum of 100 being the best a team can play. Every year, no matter how good the Duke team is, the team inevitably has a couple of "clunkers" (which for these purposes I've defined as a game score of 75 or worse). This eight minute segment against Wake represents as close to a clunker as the 2025 Blue Devils have come, but we've only played 19 games. So I compared our game score performance against the first 19 games of past Duke teams (back to 2008, which is the earliest Torvik has on his website):

YearAvgMedian<90<=75Low<=75 in games 20+NCAAT
2025​
95.89​
97​
1​
0​
87​
????
2008​
95.84​
98​
3​
0​
81​
2
16​
2009​
95.32​
98​
2​
1​
73​
3
16​
2019​
95.21​
98​
3​
0​
79​
2*
8​
2011​
94.84​
96​
2​
0​
80​
2
16​
2010​
93.58​
97​
3​
1​
48​
1
1​
2015​
91.58​
97​
5​
2​
48​
2
1​
2014​
91.42​
96​
5​
2​
70​
2
64​
2013​
91.26​
96​
3​
1​
29​
3
8​
2012​
90.47​
92​
7​
1​
59​
4
64​
2016​
90.00​
93​
7​
1​
75​
2
16​
2018​
89.95​
95​
7​
4​
63​
1
8​
2020​
89.89​
97​
7​
3​
64​
3n/a
2017​
89.63​
93​
5​
2​
40​
1
32​
2022​
89.58​
90​
7​
2​
68​
2
4​
2024​
83.11​
92​
8​
5​
40​
3
8​
2023​
78.47​
88​
11​
4​
9​
5
32​
2021​
75.79​
76​
11​
9​
50​
2n/a

* (both of the 2019 team's clunkers came while Zion was injured)

As you can see, this year's model has the best average game score, the fewest games under 90, the highest low score, and is one of only four Duke teams in the period to have no clunkers. For the first 19 games anyway. That shows not only how high a level we've played so far, but how consistent we've been (notwithstanding the first 8 minutes of the second half against Wake).

That said, Duke has had at least two clunkers from the 20th game and beyond in 14 of the previous 17 seasons (all but 2010, 2017, and 2018). Can we avoid that this year? Only time can tell.
Really cool analysis. Thing that struck me most was the lowest four averages were in the last four years!
 
Back
Top